Autism is a behaviorally defined but biologically heterogeneous disorder1 involving a cluster of impairments in language, social interaction and behavior. Large brains are common in autism, but their underlying tissue microstructure and their impact on the autistic phenotype are not understood. This proposal is based on the model that increased white matter volume is related to impaired connectivity and to underlying core processing abnormalities in autism. To test this model we will pursue our prior findings that white matter enlargement drives large brains and is localized to the later-rhyelinating radiate zone (the white matter closest to the cerebral cortex). We will use multispectral MRI to gain systematic information about the tissue microstructure and patterns of structural connectivity of enlarged white matter in autism, and we will relate these findings to cognitive and behavioral phenotype and to genome data. Forty autistic and forty typically developing boys ages 6-10 will receive a cognitive and behavioral battery and genetic studies. Specific Aim 1: We will construct a convergent multimodal tissue microstructure profile, coregistering boundaries from whole-brain segmentation and parcellation with estimated tissue parameters (T1, T2*, PD) magnetic resonance and diffusion tensor (DTI) data. Data will be analyzed for patterns discernable in multiple measures that may illuminate altered tissue microstructure. Specific Aim 2: We will investigate structural connectivity utilizing DTI tract segmentation and multimodal tissue profiling of segmented tracts. Specific Aim 3: We will test brain-connectivity-processing-endophenotype correlations, including genome data. We predict 1) that perturbations of volume and tissue integrity will be correlated with each other, and will be distributed by zone-more in later-myelinatmg radiate than deep zone white matter[unreadable]rather than by tract or neural system;2) thatmeasures of impaired complex processing (e.g., central coherence) will a) correlate with measures.of widespread anatomical changes (e.g., white matter increase), b) be associated with autism even if regional brain changes are absent, and c) be associated in severity with.severity of specific behavioral endophenotypes as well as of widespread anatomical abnormalities, and 3) that common anatomical and behavioral findings will be associated with varied genome changes. Our study design will also detect alternate outcomes and will yield data relevant for multiple levels of translational research.